Danish national guideline for the treatment of Clostridioides difficile infection and use of faecal microbiota transplantation (FMT)

Abstract

Aim: This Danish national guideline describes the treatment of adult patients with Clostridioides (formerly Clostridium) difficile (CD) infection and the use of faecal microbiota transplantation (FMT). It suggests minimum standard for implementing an FMT service.

Method: Four scientific societies appointed members for a working group which conducted a systematic literature review and agreed on the text and recommendations. All clinical recommendations were evalluated for evidence level and grade of recommendation.

Results: In CD infection, the use of marketed and experimental antibiotics as well as microbiota-based therapies including FMT are described. An algorithm for evaluating treatment effect is suggested. The organisation of FMT, donor recruitment and screening, laboratory preparation, clinical application and follow-up are described.

Conclusion: Updated evidence for the treatment of CD infection and the use of FMT is provided.

Keywords:

• Clostridioides difficile
• fecal microbiota transplantation

Introduction

Clostridioides (formerly Clostridium) difficile (CD) infection (CDI) is a major cause of nosocomial diarrhoea and accounts for 20–30% of cases of antibiotic-associated diarrhoea [1]. The disease poses a persistent health threat, is associated with a high mortality and generates considerable hospital costs [2–5].

Faecal microbiota transplantation (FMT) is the transfer of minimally processed faeces from a healthy donor to a patient in order to treat disease [6]. The method has been used in modern medical science since 1958 [7]. Its clinical effect in recurrent CDI (rCDI) has been documented in observational [8–13] and randomised [14–19] studies. The use of FMT in the treatment of other conditions is being explored in clinical trials. In the future, the use of microbiota-based drugs may potentially replace or supplement FMT. The basis for this clinical guideline is the use of FMT in patients with CDI. Experimental treatments and indications for FMT are briefly discussed.

This Danish national guideline describes the treatment of adult patients with CDI and the use of FMT.

Methods

Guideline development process

The formation of this Danish national guideline followed the proforma for clinical guidelines under the Danish Society for Gastroenterology and Hepatology, including representation in the working group by specialists, doctors in training, university hospitals, regional hospitals, and all geographic regions of Denmark. The process and the final guideline were endorsed by the Danish Society for Gastroenterology and Hepatology, the Danish Society of Infectious Diseases, the Danish Society for Clinical Microbiology, and the Danish Immunology Society for Clinical Immunology, following a hearing process in each scientific society. Each scientific society appointed at least two working group members.

Definitions

Clostridioides difficile infection (CDI)

Presence of diarrhoea AND a stool sample that is positive for Clostridioides difficile (CD) toxin or toxin gene (Figure 1); in special cases of CDI, also defined as endoscopic/histopathological finding of pseudomembranous colitis. For information about which test is used to detect CD, please consult local guidelines.

Diarrhoea is defined as three or more bowel movements per day with a consistency corresponding to the Bristol Stool Form Scale (Bristol) 6-7.

Figure 1. Classification of Clostridioides difficile infection (CDI).

First CDI

First detection of CDI. Synonymous with initial/index CDI.

Recurrence of CDI

A new case of CDI occurring after discontinuation of treatment for previous CDI cases where clinical resolution was achieved during treatment (absence of diarrhoea).

Mild/moderate CDI

CDI cases not meeting the criteria for being severe or fulminant, see below.

Severe CDI

CDI with the presence of one or more signs of severe colitis (systemic impact): clinic (abdominal pain), biochemistry (hypoalbuminemia (p-albumin < 30 g/l); leukocytosis (leukocytes > 15 × 10⁹/l); leukopenia (leukocytes <2 × 10⁹/l); renal impairment (creatinine above 1.5 × premorbid level).

Fulminant CDI

Severe CDI and one or more of the following: clinic(cognitive affection; fever above 38.5 °C; ileus; hypotension/septic shock), course (development of multi-organ failure; need for intensive therapy), imaging (ileus; toxic megacolon (colon diameter> 6 cm)), endoscopy (pseudomembranous colitis).

Treatment-refractory CDI

Persistent diarrhoea after five days of anti-CD antibiotic treatment, or worsening at any time during treatment. Cessation of diarrhoea is defined by daily bowel movements <3 and Bristol <6. Any differential diagnoses should be considered in the absence of treatment effect. The term treatment refractory is used synonymously with treatment failure, which, however, is used by some in the sense of recurrence and is therefore avoided in this guideline.

Faecal microbiota transplantation (FMT)

Transfer of minimally processed donor faeces from a donor to a patient in order to establish normal bowel function.

Resolution of CDI

Absence of CD-related diarrhoea after conclusion of treatment, i.e. either a maximum of three non-liquid stools (Bristol 5 or lower) daily or, in case of diarrhoea, a negative CD test.

Level of evidence and grade of recommendation

Assessments of evidence level (EL 1–5) and grade of recommendation (RG A-D) follow recommendations established by the Centre for Evidence-Based Medicine, University of Oxford.

Literature search method

The National Library of Medicine, Pubmed, and The Cochrane search databases were used. Free text search was employed, and the following MeSH term was used; fecal microbiota transplantation. The following search string was used to identify clinical trials using FMT in CDI: (fecal OR faecal) AND ((microbiota AND (therapy OR installation OR transplant OR transplant) OR bacterotherapy) OR stool transplant) AND (CDI OR rCDI OR Clostridioides difficile OR Clostridium difficile OR Clostridium infection OR Recurrent Clostridium difficile infection) AND (colonoscopy OR oral OR capsule OR capsules OR nasojejunal tube OR nasoduodenal tube OR nasogastric tube OR enema OR retention enema OR efficacy OR retrospective OR experience). The literature search was concluded on 22 April 2020.

Treatment of C. difficile infection

Metronidazole and vancomycin

Two small non-blinded, randomised trials from 1983 (n = 88) and 1996 (n = 62) compared treatment with metronidazole and vancomycin and found no difference in efficacy (cure rate) within 3–4 weeks [20,21]. Based on these studies and due to the large price difference, the US guidelines from 1997 in place at the time of the studies recommended metronidazole as first-choice treatment [22]. In 2007, the first randomised blinded study stratified by CDI severity in a study including 172 patients and follow-up after three weeks showed a significantly higher cure rate of vancomycin of 97 versus 84% after metronidazole [23]. The difference in effect was observed in the group with severe disease. In another study with data collected from two multinational studies, vancomycin (n = 266) was also superior to metronidazole (n = 289) with an 81 against a 73% cure rate [24]. Furthermore, a retrospective cohort study from 2016 found that 30-day mortality and recurrence risk (evaluated up to 56 days) were lower after vancomycin than after metronidazole [25]. Furthermore, if vancomycin is minimally absorbed from the gastrointestinal tract, the concentration of active substance in the intestine is significantly higher than with metronidazole. The latter is absorbed to a greater extent, which is associated with a risk of several systemic side effects.

In severe or fulminant CDI, increasing the dose of vancomycin from 125 mg × 4 to 500 mg × 4 has been considered. Two studies examined the effect of different doses, finding no effect of a dose increase [26,27]. In contrast, it has been proposed that a dose increase may have a negative effect on the intestinal microbiome.

In fulminant CDI, the addition of intravenous metronidazole to oral vancomycin in a single-centre retrospective study of 88 patients in an intensive care unit showed reduced mortality [28]. However, in a recent two-centre retrospective study including526 patients with CDI of varying severity including fulminant CDI, the addition of intravenous metronidazole had no effect on death, colectomy or recurrence [29]. It is generally agreed that in case of paralytic ileus or threatening toxic megacolon, the addition of intravenous metronidazole to oral or rectal vancomycin improves antibacterial treatment and protection against other anaerobic bacteria in the bloodstream. Intravenous metronidazole cannot be used as monotherapy for the treatment of CDI. If the patient is unable to take vancomycin capsules, administering vancomycin in liquid form (oral solution or infusion solution) is recommended, either by drinking it, through a nasointestinal tube, or rectally by endoscopy.

More than 95% of intravenously administered vancomycin is excreted unchanged through the kidneys. Only a very limited amount is excreted in bile and faeces, so intravenous vancomycin prescribed on another indication has no effect on CDI, and oral or enteral vancomycin must be supplemented according to the above principles for concomitant CDI.

Fidaxomicin and vancomycin

Fidaxomicin is a narrow-spectrum antibiotic against gram-positive aerobic and anaerobic bacteria, which is believed to be microbiotasparing compared with vancomycin. Like vancomycin, it is minimally absorbed from the gastrointestinal tract. In 2011 and 2012, two phase-3 studies were published with 629 and 535 patients, respectively. Both studies showed the same treatment effect for vancomycin and fidaxomicin in relation to treatment of the current infection, but fidaxomicin significantly reduced the recurrence rate compared with vancomycin (vancomycin 25–27% versus fidaxomicin 13–15%) [30,31]. Similar findings were described in a study where fidaxomicin for five days followed by pulse treatment every other day until day 25 was compared with standard vancomycin for ten days [32]. Most trial participants in the first two trials were treated for their first CDI, whereas only a small subset had recurrent CDI. In this group, a similar difference was seen between the recurrence rate of fidaxomicin- and vancomycin-treated patients, but the difference observed in this subgroup was not significant. In a randomized, 3-armed, clinical trial including recurrent CDI – FMT (n = 24), fidaxomicin (n = 24) for ten days or vancomycin (n = 16) for ten days – the difference between fidaxomicin (cure rate 42%) and vancomycin (cure rate 19%) was not statistically significant with cure rate defined as absence of diarrhoea recurrence within eight weeks [19]. Both antibiotics were significantly poorer than FMT, which showed a 92% cure rate. Fidaxomicin has not been compared with FMT in other studies.

Studies on and experience with fidaxomicin treatment in patients with recurrent CDIs, including multiple recurrences, are generally lacking.

Vancomycin pulse-tapering therapy

In recurrent CDI, prolonged treatment with vancomycin with dose reduction or pulse therapy has been studied in few studies. In a retrospective study from 2017, 100 patients with an average of three relapses were treated with vancomycin 125 mg four times daily for ten days, followed by vancomycin 125 mg once daily for one week and then either a 125 mg dose every other or every three days, the latter producing a prolonged regime [33]. Calculated after 90 days, the treatment effect was superior (cure rate 81%) in those who received treatment every three days compared with those who received treatment every other day (cure rate 61%). In a series of 22 patients, most of whom had three or more recurrences, the vancomycin 125% × 4 daily cure rate was reduced to 125 mg × 2 daily for one week, 125 mg × 1 daily for one week and then pulse treatment with 125 mg every two days for a week and lastly pulse treatment with 125 mg every three days for one week [34]. Conversely, a propensity score-matched cohort study with patients with first or second recurrence of CDI found no difference in recurrence rate after vancomycin with (n = 226) or without tapering (n = 678). However, they did observe a difference in mortality after 90 days in favour of the downsizing regime, but not after 180 days [35]. No current randomised trials have investigated different tapered or pulse regimens for vancomycin, but one large, double-blind study is ongoing with standard fidaxomicin for vancomycin decoupling over 31 days (Clinicaltrials.gov registration number NCT02667418). In one study, tapering with vancomycin seemed superior to a single enema FMT, whereas colonoscopic FMT was three times better than vancomycin [15].

Studies of long-term treatment with tapering or pulse treatment were summarized in a recent review [36]. This review also described studies in which vancomycin treatment was combined with probiotics and rifaximin.

Comparisons of metronidazole, vancomycin and fidaxomicin were summarized in a network meta-analysis from 2020. In this meta-analysis, vancomycin and fidaxomicin were effective first-line treatments for mild/moderate CDI as opposed to metronidazole, and fidaxomicin was possibly superior to vancomycin in avoiding CDI recurrence [37]. In addition, a number of overviews and guidelines review the established treatments and new treatments and provide recommendations from the clinical societies SHEA/IDSA, ESCMID, ASID and ACG [38–42].

Other CDI-active antibiotics

In addition to metronidazole, vancomycin and fidaxomicin, a number of other antibiotics are also active against CD. Although the above mentioned three antibiotics predominantly cover the treatment need, the following may be considered, e.g. in the case of significant side effects, allergies or interactions related to the first three.

Nitazoxanide

Nitazoxanide is primarily a anti-parasite drug, but in a single study it has also shown activity against CD. Musher et al. [43] described the results of a small, double-blind, randomised study including 49 patients with CDI who were treated with either vancomycin 125 mg × 4 or nitazoxanide 500 mg × 2, both for ten days. The study did not have the strength to show non-inferiority for nitazoxanide relative to vancomycin but did have comparable results: The response rate was 73% (16/22) in the nitazoxanide group and 67% (18/27) in the vancomycin group (RR 1.09, 95% CI 0.75–1.58). Nitazoxanide is not marketed in Denmark, but may be ordered from a pharmacy that holds a dispensing permit from the Danish Medicines Agency.

Rifaximin

Rifaximin is an orally ingested antibiotic that is poorly absorbed; it belongs to the rifamycin group. Rifaximin is active against CD, but resistance may occur with ribotypes 027/ST1 and 017/ST37 (Unpublished). In a clinical, randomised study from 1990 with a total of 20 patients, Boero [44] found no significant difference in outcome after vancomycin or rifaximin (RR 0.90, 95% CI 0.69–1.18). In a systematic review from 2018 [45], eight studies from 2007–2017 were reviewed, including two randomised studies. One common finding was that rifaximin was used primarily as a relapse prophylaxis immediately after stopping either vancomycin or metronidazole treatment. In a random-effects meta-analysis of the two randomised clinical trials (RCT) included in the meta-analysis, the effect of rifaximin, although not statistically significant, was compared with either placebo or metronidazole when given as recurrence prophylaxis (OR 0.36 (95% CI 0.12–1.10)).

Teicoplanin

Teicoplanin is usually used intravenously. In oral administration, it is absorbed only to a very modest extent from the gastrointestinal tract. Nelson’s 2017 Cochrane review [39] used data from two smaller randomised trials [21,46] that compared teicoplanin withvancomycin. The success rate was higher in the teicoplanin group (87%; 48/55) than in the vancomycin group (73%; 40/55) (relative risk (RR) 1.21, 95% CI 95% 1.00–1.46). In both studies, the duration of treatment was ten days and the dose of vancomycin was 500 mg x 4. In de Lalla’s study, the dose of teicoplanin was 100 mg × 2 and in the Wenish study, the dose was 400 mg × 2. A Serbian study [47] from 2018 compared the effects of teicoplanin to those of vancomycin. Patients with severe CDI were treated with teicoplanin at 100 mg × 2 versus vancomycin 125 mg × 4 for ten days. Patients with severe fulminant CDI were all initially given metronidazole 500 mg × 3 IV together with either teicoplanin 200 mg × 2 or vancomycin 500 mg × 4. The overall clinical cure rate for both groups of severity was higher in the teicoplanin group (97/107; 91%) than in the vancomycin group (143/180; 79%) (odds ratio (OR) 2.51 CI 95% 1.19–5.28). Likewise, clinical cure was higher for both groups separately, but this was not statistically significant. The recurrence rate was significantly lower in patients treated with teicoplanin (9/97; 9.3%) than in patients treated with vancomycin(49/143; 34.3%), p<.001, OR (95% CI) 0.20 (0.09–0.42)). Experience with teicoplanin as a treatment for CDI in Denmark is limited, partly due to the significantly higher price compared with vancomycin.

Bezlotoxumab

Bezlotoxumab is a human monoclonal antibody that binds toxin B produced by CD with high affinity, thereby neutralizing the toxin. Bezlotoxumab is administered intravenously as a single dose and is used as an adjunct to ongoing antibiotic therapy (metronidazole, vancomycin or fidaxomicin). Bezlotoxumab is not a treatment for CDI and has no effect on the current CDI episode, but may be used prophylactically for new CDIs as it provides passive immunity to toxins produced by persistent or newly acquired CD spores. The applicability of bezlotoxumab in clinical practice is limited by a high cost. In two large double-blind multinational phase 3 RCTs, a significant reduction in the recurrence rate was seen after 12 weeks by comparison between bezlotoxumab and placebo given in addition to standard antibiotic treatment (16–17% recurrence rate versus 26–28% in the placebo group, p<.001) [48]. In these studies, the number needed to treat (NNT) was found to be ten; and for the subgroup of patients over 65 years of age or CDI cases within six months, the NNT wassix. Bezlotoxumab has been approved in adults for adjuvant treatment in this particular subgroup. However, experience and evidence concerning prophylactic use of bezlotoxumab against rCDI or as a CDI treatment option are generally lacking. Specifically, knowledge is missing that may define the place of treatment within the treatment algorithm.

Oral vancomycin prophylaxis (OVP) against new CDI by other concomitant antibiotic action

In retrospective case series, a reduced risk of recurrent CDI has been found when vancomycin is given concomitantly with antibiotic treatment for other infections. The studies compared patients who received OVP with patients who did not receive OVP (non-OVP) during other antibiotic treatment. A survey comprising 551 patients was found that OVP (125 mg × 4) reduced the risk of CDI recurrence in patients older than 65 years with two previous CDIs and where OVP was given for more than 50% of the time that another antibiotic treatment was given [49]. In another study counting 203 participants, patients had received OVP (n = 71, such as 125 mg × 2 or 250 mg × 2) for up to one week during other concomitant antibiotic treatment. Here, a reduction in the recurrencerate was seen (27% in the non-OVP group against only 4% in the OVP group, p<.001) [50]. The studies were described in a review [51], which also provided information about the prospective studies to clarify doses and treatment durations, which were reported to www.clinicaltrials.gov.

Long-term or lifelong vancomycin treatment

For patients with persistent, recurrent CDI, where other treatment options seem exhausted or are contraindicated or where the patient does not want, e.g. FMT, empirical long-term treatment with vancomycin is used in the lowest possible dose, possibly for years/lifespan in case the expected remaining life is short, i.e. as secondary prophylaxis also without exposure to other antibiotics. This practice has been sparsely researched and described. A retrospective case series described the use of long-term treatment with vancomycin (defined at 125 mg once daily for at least eight weeks after tapering over 14 days from 125 mg × 4 daily for 14 days) in a group of 20 elderly patients with multiple previous recurrences who were either not candidates for FMT or had not benefitted from multiple FMT treatments. The patients were not exposed to other antibiotics during the period. Only one patient experienced recurrence during long-term treatment. This case was successfully treated with FMT. Some patients were treated for life/during the entire follow-up period (up to 19 months). Recurrence was seen in 31% of patients in whom vancomycin was later discontinued. No side effects were observed. Four patients died during follow-up, but none due to CDI [52].

Intestinal microbiota treatment

In light of the increased incidence of CDIs and with increasing challenges associated with recurrent CDIs over the next 10–15 years, attempts to restore a normal intestinal microbiota using intestinal microbiota treatments have attracted much attention. Both nationally and internationally, the focus has been on FMT.

FMT is the transfer of faeces from a healthy donor to a patient in order to treat disease. FMT is reviewed in this guideline in a separate section, and here only the role of FMT in the treatment of patients with CDI is summarised.

Another approach to restoring normal intestinal microbiota is instillation of a standardised bacterial culture also coined rectal bacterial instillation (RBI)/rectal bacteriotherapy (RBT). RBI differs from FMT in that it does not take into account donor recruitment and screening and it eliminates the, albeit modest, risk of transmitting unknown pathogens. The method is not widespread internationally, but has been used on a smaller scale in Denmark since the 1980s and is therefore also discussed below.

FMT in recurrent CDI

FMT has been described as an effective treatment for recurrent CDI in case series since 1958 [7]. However, the first RCT documenting the beneficial effect compared with standard vancomycin treatment was not published until 2013 [14]. Here, the authors compared FMT administered via a nasoduodenal tube (preceded by 4–5 days of high-dose vancomycin) with oral vancomycin alone and oral vancomycin followed by cleansing. After ten weeks, 81% were non-recurrent in the FMT group against only 31 and 23% in the two vancomycin groups, respectively (p=.008/.003). Two subsequent RCTs have shown a similar effect for FMT administered via colonoscopy over 1–4 times (after three days of vancomycin) versus oral vancomycin for 10 days followed by pulse treatment for a minimum of three weeks (90 versus 26%, p<.0001) [15] and for FMT administered via colonoscopy or naso-jejunal tube (preceded by 4–10 days of vancomycin) versus fidaxomicin for ten days or vancomycin for ten days (92 versus 42 versus 19%, p=.0002/<.0001) [19]. The latter study also showed that FMT was superior to fidaxomicin in preventing new recurrences. A single study was unable to demonstrate the same effect as the others, as no difference in effect was found after 120 days of follow-up between a single FMT given via rectal installation, enema, and oral vancomycin for 14 days followed by tapering over four weeks. Use of single enema administration may explain non-superiority [53].

In American and European guidelines, FMT is now recommended for patients with recurrent CDI who have not had an effect of repeated antibiotic regimens. No clear recommendations exist for when FMT should be offered. The published RCTs have recruited patients with a median of 3–4 CDI relapses, so evidence for the effect of FMT is strongest for patients with multiple relapses. A recent Danish RCT (not yet published) found a total effect of 1–3 rectal FMT infusions of 76%. Subgroup analysis showed that 90% of patients with first recurrence were cured, whereas the proportion was 70% among patients with second recurrence. The results for the limited number of patients in the study with multiple relapses were more mixed. These findings may strengthen an argument for using FMT earlier [54].

FMT for a patient’s first CDI

The above studies examined FMT as a treatment modality exclusively in patients with recurrent CDI. A small proof of concept RCT from Norway randomised 21 adults with first CDI (no previous CDI case) to either ten days of metronidazole or FMT via enema/cleavage. Their results indicated a possible beneficial effect of FMT for this patient group [55]. The same research group is now in the process of conducting a major RCT on the effect of FMT versus vancomycin, this time using regular donor stool. (Clinicaltrials.gov registration number NCT03796650).

Another small randomised pilot study (n = 16) compared vancomycin and donor faeces mixed from multiple donors to hospitalized patients with first CDI [56]. The study found that 8/9 patients (89%) in the vancomycin group achieved symptom resolution within three days, whereas only 4/7 (57%) patients achieved resolution after the first FMT (57%) and 5/7 (71%) patients achieved resolution after the second FMT (71%). These did not record vancomycin pre-treatment before FMT.

At the end of the literature search, no published results from randomised studies reported on the effect of FMT on the first CDI.

FMT in treatment-refractory/severe/fulminant CDI

FMT can have a beneficial effect in CDI when CDI is medically refractory or develops into severe/fulminant infection during treatment. In 2015, a cohort study described 29 patients with severe or fulminant CDI who were treatment refractory, i.e. without effect of five days of vancomycin treatment. Patients received colonoscopic FMT with repeated FMT every third, fourth or fifth day if CDI symptoms persisted [57]. Overall, 93% (27/29) of patients were cured, of whom all (10/10) had severe CDI and 89% (17/19) fulminant CDI. A single FMT was given to 18 (62%), two FMTs to nine (31%) and three FMTs to two (9%) of the patients.

An RCT from 2018 compared a single versus two or more FMT treatments in 56 admitted CDI patients (12% with their first CDI) with severe/complicated treatment-refractory (for vancomycin or fidaxomicin) disease. Here, the cure rate was 75% after one FMT (21/28) and 100% (28/28) after two or more FMT treatments. Thirty-three patients (59%) had severe-complicated (fulminant) CDI (17 in the single FMT group; 16 in the multiple FMT group) and 36 patients (64%) had pseudomembranous colitis (17 in the single FMT group; 19 in the multiple FMT group).

In a retrospective cohort study with analysis of three-month mortality in 111 patients admitted with CDI (66 in the FMT group and 45 in the non-FMT group), a pronounced reduction in mortality was found after FMT [58]. Mortality was 24% (27/11) three months after diagnosis of CDI among which 12% (8/66) occurred in the FMT group versus 42% (19/45) in the non-FMT group. NNT was only two for patients with severe CDI who needed FMT treatment to save one life. A similar NNT of three for FMT to prevent death from severe/fulminant CDI was found in a retrospective, matched cohort study of 48 CDI patients admitted to the intensive care unit (including 15 patients (31%) with severe CDI and 33 (69%) with fulminant CDI), among whom 16 received FMT and 32 non-FMT [59].

In a subgroup analysis of clinical trials before and after FMT implementation, comprising 199 patients with fulminant CDI and 110 patients with treatment-refractory (i.e. without improvement after five days of anti-CDI antibiotic treatment) severe/fulminant CDI, the investigators reported a significantly lower CDI-related mortality and colectomy rate after introduction of FMT in relation to earlier. For fulminant CDI, mortality was 21% before FMT and 9% after FMT; for refractory severe/fulminant CDI, it was 42% before FMT versus 12% after FMT. Implementation of FMT in treatment produced a significant reduction in the colectomy rate from 16% before FMT to 6% after FMT in fulminant CDI and from 32% before FMT to 8% after FMT was used in refractory severe/fulminant CDI [60].

RBI in recurrent CDI

Internationally, only one Canadian study has been published using a bacterial culture rather than donor faeces. In this study, two patients with rCDI were treated successfully [61]. In Denmark, the bacterial culture for RBI was originally developed in the 1980s at the Copenhagen University Hospital (Rigshospitalet) by Tvede et al. [62]. Until 2014, the bacterial culture was produced on a smaller scale at Rigshospitalet and used for a total of around 200 patients. In 2016, the bacterial culture was approved by the Danish Medicines Agency as a drug for use in recurrent rCDI. Today, the bacterial culture is manufactured on a larger scale. The original work [62] included six patients who had been successfully treated for rCDI. In a case series of 55 patients treated with RBI in the 2000–2012 period, clinical success was recorded for 35 (64%). In the subgroup of patients who had their first CDI episode <6 months before RBI, the success rate was 75% (25 of 32 patients) [63]. In a Danish RCT completed in March 2019 and currently only published in abstract form, patients with rCDI were randomised to three arms: Vancomycin + FMT, vancomycin + RBI, or vancomycin monotherapy with tapering. The study included a total of 96 patients. In the RBI group, 16/31 (52%) were cured against 19/34 (56%) in the FMT group (OR 1.2 95% CI 0.4–3.2) and 14/31 (45%) in the vancomycin group. The RBI treatment consisted of three treatments for three consecutive days, whereas in the study and following clinical assessment of the treatment effect, it was possible to repeat FMT up to two times with a different donor within 14 days after the first FMT. In FMT given 1–3 times (24 patients received one, seven received two and one received three FTM) within 14 days, FMT was significantly superior to RBI, with success rates of 26/34 (76%) and 16/31 (52%), respectively (OR 3.0, 95% CI 1.1–8.8). RBI and vancomycin were equally effective (p = .61) [54].

When assessing the overall clinical evidence for RBI, no clinical recommendation can be made for its use.

Surgery for CDI

In fulminant refractory CDI, mortality can reach 80–90% [64]. Patients with fulminant CDI should therefore undergo close multidisciplinary monitoring and assessment in the absence of a treatment response for surgical intervention.

Colectomy

Surgical intervention with colectomy appears to be able to improve survival in fulminant CDI. In a study of 165 patients with fulminant CDI admitted to the intensive care unit, the mortality rate with continued medical treatment was found to be 65%, whereas the mortality was 36% after colectomy in patients over 65 years [65]. It is well documented that timely surgery may save lives, but no objective clinical and paraclinical parameters currently predict the course of CDI after starting treatment with antibiotics or FMT. With premature surgery, the possibility of healing through medical treatment alone is lost, and the patient is subjected to surgical risk and ileostomy. In the event of late surgery, the risk of postoperative complications and death increases.

Loop ileostomy

An alternative to total colectomy or subtotal colectomy with blind closure of the rectum is the presentation of a loop ileostomy with the possibility of antegrade lavage with fluid, vancomycin or possibly FMT. The procedure of laparoscopic loop ileostomy, perioperative rinsing with fluid followed by rinsing with vancomycin is not inferior to colectomy in terms of mortality [66–72]. To date, antegrade lavage with donor faeces has not been described in the literature. Loop ileostomy is thus a recognized option but should be chosen over colectomy only in the absence of signs of perforation or toxic megacolon.

Endoscopic desufflation and retrograde installation of vancomycin or FMT

May be considered in patients with colon dilatation and without upper access, but is not recommended in patients with fulminant CDI.

Assessment of treatment effect in CDI

A treatment algorithm for CDI is presented in Table 1. The treatment effect is assessed continuously based on the clinical picture (Table 2). Clinical recommendations including assessment of evidence level and grade of recommendation are presented in Table 3.

Table 1. Treatment algorithm in patients with Clostridioides difficile infection (CDI).

CDI classification	Recommended treatment
All cases of CDI	Revise ongoing antibiotics and proton pump inhibitors
First CDI	Mild/moderate: First choice: Oral vancomycin 125 mg × 4 for ten days Second choice: If vancomycin is not available (non-hospitalised): Oral metronidazole 500 mg × 3 for ten days Severe and/or inpatient: First choice: Oral vancomycin 125 mg × 4 for ten days Second choice: Oral fidaxomicin 200 mg × 2 for ten days Consider faecal microbiota transplantation (FMT)
Recurrent CDI: First recurrence	Mild/moderate (not requiring hospitalisation): First choice: Oral vancomycin 125 mg × 4 for ten days FMT may be considered Severe and/or inpatient: First choice: Oral vancomycin p.o. 125 mg × 4 for ten days Second choice: Oral fidaxomicin 200 mg × 2 for ten days FMT should be offered as an extension of antibiotic treatment*
Recurrent CDI: Second recurrence and above	All degrees of disease: First choice: FMT should be offered in continuation of antibiotic treatment* If FMT cannot be completed: First choice: Oral vancomycin. A tapering regimen may. Be applied, e.g.: 125 mg × 4 daily for two weeks, 125 mg × 2 daily for one week, 125 mg × 1 daily for one week, 125 mg × 1 every two days for one week 125 mg × 1 every three days for two weeks
Fulminant CDI (regardless of number of episodes)	All: Multidisciplinary** assessment should be made with regard to treatment plan. First choice: FMT (possibly repeated at 3-day intervals) Antibiotics: Intravenous metronidazole 1,000 mg × 1 or 500 mg × 3 daily in addition to oral/rectal vancomycin Surgery considered: Colonoscopic desuffling with vancomycin flushing/FMT, loop ileostomy with vancomycin flushing/FMT – or colectomy.
Refractory CDI (regardless of number of episodes)	Revision of antibiotic treatment, both CDI-related and non-CDI-related First choice: FMT (possibly repeated at 3-day intervals) Consider multidisciplinary assessment**

*FMT is preceded by 4–10 days of oral vancomycin (/fidaxomycin) with discontinuation 1–2 days before FMT. With a longer waiting time for FMT, vancomycin/fidaxomycin is continued, possibly with tapering to the lowest possible dose that provides symptom relief. Thus, when referring to FMT, no end date is set for antibiotic treatment.

**Multidisciplinary assessment: gastroenterology, possibly infectious disease medicine, colorectal surgery, clinical microbiology.

Table 2. Evaluation of treatment effects in patients with Clostridioides difficile infection (CDI).

Definition of diarrhoea: ≥3 liquid stools per day (Bristol score 6–7), assessed over at least two days. CD test: PCR-based toxin test or faecal culture for Clostridioides difficile. CD testing is performed in clinical routine only in patients with diarrhoea and not during ongoing antibiotic treatment.
A] Clinical evaluation during treatment
Continued or worsened diarrhoea ≥3 liquid stools per day (Bristol 6–7) Failure Refractory CDI, change treatment as shown in Table 1	Clinical improvement < 3 stools per day, no or decreased pain Effect Complete treatment
B] General evaluation following treatment
<3 stools per day: no diarrhoea Effect CD test is not recommended (if a CD test has been performed and is positive, this indicates carrier status without clinical disease)
New-onset diarrhoea: ≥3 liquid stools per day (Bristol 6–7) and a positive CD test: Failure (CDI recurrence)	New-onset diarrhoea: ≥3 liquid stools per day (Bristol 6–7) and a negative CD test: Effect (Diarrhoea of other cause)

Table 3. Clinical recommendations for the treatment of patients with Clostridioides difficile infection (CDI).

Clinical recommendations for the treatment of patients with Clostridioides difficile infection (CDI)

If necessary, discontinue antibiotic therapy, in particular cephalosporins, clindamycin and fluorquinolone, which may trigger CDI (EL 1a, RG A). The first mild-to-moderate severity CDI is treated with vancomycin 125 mg orally × 4 for ten days. In mild initial cases, in the unaffected patient seen in the primary sector and where vancomycin is unavailable, metronidazole 500 mg orally × 3 for ten days is used (EL 1a, RG A). Vancomycin is dispensed free of charge from the country’s hospitals. Please see regional guidelines and guidelines from the Danish Regions. Vancomycin is superior to metronidazole in terms of efficacy, recurrence rate and mortality in CDI (EL 1a, RG A). First cases of severe CDI (or mild-to-moderate CDI in a hospital) are treated with vancomycin 125 mg orally × 4 for ten days (first choice) or fidaxomicin 200 mg orally × 2 for ten days (EL 1a, RG A). In case of CDI recurrence, vancomycin 125 mg orally x 4 for ten days (first choice) or fidaxomicin 200 mg orally × 2 for ten days (EL 3a, RG B). At the first recurrence of CDI, FMT can be offered (EL 1b, RG A). Patients with two or more recurrences of CDI should be offered FMT (EL 1a, RG A). In case of multiple CDI recurrences, where FMT is not indicated or cannot be performed, vancomycin tapering may be tried, for example 125 mg four times daily for two weeks, 125 mg twice daily for one week, 125 mg once daily for one week , 125 mg once every two days for one week, 125 mg once every three days and then discontinuation, corresponding to a total of seven weeks of treatment (EL 3b, RG B). No clinical recommendation can be given for the use of rectal bacterial installation (RBI) (EL 1b, RG A). In the case of fulminant CDI with paralytic ileus or toxic megacolon where a threatening surgical indication is considered: metronidazole 1,000 mg IV x 1 or 500 mg IV x 3 daily in addition to vancomycin (EL 3b, RG C). In case of severe or fulminant CDI, the patient should undergo multidisciplinary evaluation to consider a surgical intervention (endoscopic desuffling, endoscopic rinsing with vancomycin or FMT, colectomy, loop ileostomy with perioperative rinsing and antegrade vancomycin and possibly FMT; for toxic megacolon or colon perforation, a total/subtotal colectomy is performed) (EL 3a, RG C). FMT is currently not recommended for treatment of mild-to-moderate first CDI (EL 1b, RG B). In case of treatment refractory CDI, FMT should be considered (EL 2b, RG B). In case of severe/fulminant CDI, FMT should be considered (EL 2b, RG B). Bezlotoxumab may be considered as an adjuvant treatment for antibiotics by CDI in patients who are at a high risk of recurrence of CDI and where FMT is not desired, but its place in the treatment algorithm currently remains unknown and limited by high treatment cost (EL 1b, RG B). To prevent recurrence in selected fragile patients with recurrent CDI where other treatment attempts have been unsuccessful or the patient does not want other treatment, long-term vancomycin may be used, reduced to the lowest possible dose at which the patient remains free of diarrhoea (EL 4, RG C). Prophylactic vancomycin during other antibiotic therapy may be used in elderly patients who have had at least one previous CDI. The dose is 125 mg orally twice daily until seven days after completion of other antibiotic therapy (EL 4, RG C). During waiting time for FMT, either due to a temporary relative contraindication or other antibiotic treatment, long-term vancomycin in the lowest possible dose may be used as a bridge until FMT is available or contraindications for FMT have ceased (EL 5, RG D). Treatment effect after completion of treatment is assessed by clinical resolution of diarrhoea. Only in case of renewed diarrhoea should the treatment be supplemented with a CD test (EL 2, RG B). CD testing during treatment is not recommended (EL 4, RG C).

Persistent or worsening diarrhoea during antibiotic treatment indicates refractory disease, in which case treatment is escalated (Table 1). If an effect is observed during antibiotic treatment, the total treatment effect may be assessed one week after treatment cessation. The clinical effect is primarily assessed, i.e. resolution of diarrhoea. Diarrhoea assessment is supplemented with CD testing, as approximately 2% of the healthy population have positive CD test results without experiencing clinical disease [73]. The carrier frequency is higher during hospitalisation and in cases with high comorbidity and age. It may take weeks before the toxin test becomes negative after a clinical treatment effect was observed.

CD testing during ongoing treatment with vancomycin/fidaxomixin/metronidazole cannot be used as the sample may be falsely negative due to poor excretion of toxins. Thus, testing cannot be used to distinguish between refractory disease and non-CDI/late effects of CDI [74]. This is assessed based on the clinical picture. Repetition of tests for double determination is not recommended, as the method is very sensitive [75]. Genetic testing at intervals of less than a week will generally not be meaningful.

Persistent bowel symptoms in combination with a negative CD test after treatment may be due to post-infectious irritable bowel syndrome [76,77]. In case of new-onset diarrhoea, CD testing may help to distinguish between recurrent CDI, treatment-requiring diarrhoeal disease or post-infectious irritable bowel syndrome.

Faecal microbiota transplantation (FMT) – clinical application

Treatment responsibility

Clinical application and follow-up are performed by a specialist in gastroenterology/hepatology or infectious medicine under the auspices of a medical hospital department, or by delegation from a specialist to another doctor/other professional group, and in accordance with agreed clinical guidelines.

Treatment responsibility and responsibility for clinical follow-up rests with the attending physician. Traceability for at least 30 years between donor and recipient must be ensured in accordance with the Tissue Act. Responsibility for traceability rests with the stool bank.

Indications and contraindications to FMT

The use of FMT in patients with CDI is described above, including clinical recommendations.

In ulcerative colitis, FMT has been studied in four RCTs with inconsistent results and without long-term follow-up. Therefore, FMT is currently not recommended in ulcerative colitis outside clinical trials [78–81].

In Crohn’s disease, clinical studies have not yet been performed to assess the potential effect of FMT, but a pilot study suggests a beneficial effect in patients where donor microbiota colonizes in the recipient [82]. The treatment is currently not recommended outside of clinical trials.

Irritable bowel syndrome (IBS)

Five RCTs describe conflicting results [83–88]. Divergent findings and great heterogeneity between the studies challenge the clinical application (different study populations, multiple treatments applied, varying application form, uncertainty about donor selection, unclear risk profile, occurrence of side effects, loss of effect during long-term follow-up). FMT is currently not recommended for IBS.

Carrier status with multi-resistant bacteria

Eradication of carrier condition with multi-resistant bacteria has been sporadically reported after FMT [89–92]. A randomised study did not find that FMT was better than placebo, but did not include a sufficient number of patients to confirm or disprove the null hypothesis [93]. The treatment is recommended for use in clinical trials only.

Other conditions

FMT is being investigated for a future role in the treatment of, e.g. hepatic encephalopathy [94], pouchitis [95–99] and psoriatic arthritis [100]. The clinical experiences are too few or contradictory for a treatment recommendation to be given at present. When using FMT for new indications, treatment should be initiated only after appropriate contact with the authorities.

Contraindications and precautions

No absolute contraindications to FMT have been described. FMT can be applied by rectal access in patients with ileus. Food items associated with serious food allergies, such as peanuts, should be used with caution by donors due to the risk of food allergens being transmitted. Pronounced leukopenia and other signs of severe immune incompetence should also lead to increased caution and heightened attention and may possibly place special demands on faecal suspensions.

Patient preparation before FMT

Ensuring indication

In patients with CDI, other causes of chronic diarrhoea may occur and should be investigated legeartis.

Patient guidance

FMT is a new treatment principle with unknown mechanisms of action and long-term consequences [6]. Thorough oral and written patient guidance should therefore be given before treatment. The guidance should take into account that the treatment principle is undergoing development and that aesthetic and ethical reservations may be associated with receiving the treatment. Patients should be informed of possible side effects and their frequency. Examples of written patient guidance are available as appendices.

Planning the application

It is clarified in collaboration with the patient how the treatment will be given (see below).

Patient consent

Transmission of health information to the database and medical records for long-term follow-up when the patient has completed active treatment requires written patient consent. Examples of information and declarations of consent are available as appendices.

Minimisation of risk factors

Patients with CDI are often elderly, multimorbid patients with many medical issues [1]. Risk factors for recurrence of CDI are age above 65 years, high comorbidity, use of concomitant antibiotics and use of proton pump inhibitor [101–103]. Prior to FMT it is therefore recommended to revise medicines, in particular by discontinuing antibiotics and proton pump inhibitors, as well as optimizing the patient’s nutritional and hydration status. Anaemia has been found in a small study to be associated with failure of FMT, but it remains unclear whether correction of anaemia reduces the risk of recurrence after FMT.

Antibiotic treatment before FMT for rCDI is included as part of the overall treatment. In most clinical trials on FMT, a regimen of 4-10 days of vancomycin treatment before FMT and discontinuation of vancomycin 24–48 h before FMT is practiced [40,104]. In practice, FMT therefore means FMT preceded by vancomycin 125 mg × 4 for 4–10 days. Vancomycin is discontinued 1-2 days before administration of FMT, depending on the method of application and the frequency of the patient’s bowel movements. In case of intolerance or lack of effect of vancomycin, fidaxomicin may be given as a pre-treatment. The effect of FMT without prior antibiotics is unknown and is currently being studied in clinical trials [55,105].

Cleansing before FMT depends on the application method and is not a prerequisite for the FMT application itself. Cleansing without FMT hardly has an independent effect on rCDI [14]. However, in an observational study with colonoscopically applied FMT, it was found that inadequate cleansing was related to FMT treatment failure [106]. Colonoscopy cleansing is performed according to usual guidelines.

Other medical pre-treatment

In some studies, a proton pump inhibitor, motility-promoting treatment, etc. were used as accompanying treatments. The significance of this has not been clarified. For capsule-based treatment, discontinuation of an acid pump inhibitor three days before FMT is recommended to reduce the dissolution of the acid-resistant capsules in the stomach.

Clinical application and follow-up

The application method is the single factor that influence the FMT effect rate most. FMT may be given as an infusion with rectal catheters, via colonoscopy, nasojejunal tube, gastroscopy, nasogastric tube or PEG, or as capsule treatment with glycerol-based capsules or lyophilized capsules. Colonoscopy has a high effect rate in all studies. In a meta-analysis, a 78% effect rate was reported after one application and a 98% effect rate was reported after several applications (typically a total of two) [107]. In a meta-analysis, nasogastric and nasoduodenal tube application has been shown to have equal effect in line with that of colonoscopy after one treatment (79%) and 88% after several treatments [107]. The method involves a risk of aspiration, which was observed in three patients (8%) in a Dutch observational study including 39 treatments [108]. When applied through nasojejunal tube with the tube tip distal to the Treitz’ ligament, no aspiration was observed in a single study reporting on the treatment of 11 patients with nasojejunal tube [19]. In a meta-analysis, infusion on rectal catheters had an 56%effect after a single infusion and 92after repeated (typically up to three) infusions [107]. In a randomised study with 30 patients, one infusion was not statistically significantly better than the tapered regimen with vancomycin [53]. However, in a study with 219 patients and the possibility of multiple treatments, an effect of 91% was achieved in the modified intention-to-treat population [18]. Capsules are increasingly used; in particular, a randomised study of 116 patients [109] together with observational studies suggest an effect that is in line with that of FMT by colonoscopy [110–112]. The greatest evidence for FMT still rests on colonoscopic application, which at the same time can contribute differential diagnostics. A systematic review of treatment modalities found that the different modes of administration were equivalent with the exception of rectal infusion, which was associated with a lower effect which was, however, offset by repetitive treatment [107].

Possibility of repetitive treatment

When using rectal infusion and in patients with refractory or fulminant CDI, planning repetitive treatments can be beneficial, because the effect of one FMT is lower than the effect when using other application methods of lower degrees of severity [18,53]. Primary clinical effect is evaluated after 3–7 days. In case of continued diarrhoea, FMT may be repeated after 7–14 days. In case of repetitive administration of FMT, vancomycin is not administered in the period between individual applications [15,18].

Planned serial FMT

In patients with a high risk of recurrence, multiple treatments with FMT may be effective [113]. Possible risk factors for recurrence include severe infection [106], anaemia [19] and liver cirrhosis [114]. Treatments with serial FMT can be given at shorter intervals (e.g. three days), as the decision does not depend on clinical evaluation like in repetitive treatment. The optimal treatment regimen has not been determined and the specific treatment is planned with the patient. In case of continued symptoms after FMT, testing for CD should be repeated as a negative toxin test will suggest a cause other than CDI. However, this should not delay repetition in the above cases.

Amount of donor stool per treatment

In the published studies, down to 3 grams [115] was successfully used for whole donor stools, but most commonly 30–50 grams of donor stool was used [11,116–118]. A cohort study found that effects after a low dose (7.5 grams) and a high dose (45 grams) of donor faeces were comparable [119]. A systematic review indicated a lower treatment effect when using less than 50 grams [107]. In most studies, at least 50 grams of donor faeces are used, and it is recommended that 50 grams of donor faeces be transferred at one FMT. Although a dose-response effect likely exists, a lower dose can be defended.

Post-treatment observation

The clinical significance of post-treatment observation and the duration of observation are unclear. It is recommended that patients remain in observation for 30 min after colonoscopic application lying in a right-sided position. After rectal infusion, the patient is observed lying in a left-sided position for one hour. The recommendation for fulminant CDI patients is in-hospital treatment.

Efficacy measures

All patients who have received FMT should be evaluated clinically to confirm lasting treatment effect. Impact measures may vary depending on the purpose. In randomised studies, different times of primary effect have been used, from 8 to 12 weeks. For consistent assessment of effect, this guideline recommends that clinical effect be assessed eight weeks after treatment, which is in line with international guidelines for CDI [40]. In case of diarrhoea, monitoring should be supplement with a CD test. When evaluating FMT for rCDI, a primary effect measure is proposed in the form of clinical resolution, defined as the absence of CDI-associated diarrhoea after eight weeks or as normal bowel movements (three or fewer bowel movements per day and Bristol Stool Scale score of 5 or lower), or by diarrhoea (more than three thin (Bristol 6–7) stools daily) or a negative CD test.

Recurrence after primary resolution

With recurrence of diarrhoea in a patient who has previously received FMT and experienced a clinical effect, a CD test can distinguish between CD-related disease (recurrence) and non-CD-related diarrhoea. It is therefore recommended that patients with recurrence of diarrhoea after an otherwise successful treatment be evaluated with CD testing. In case of a positive CD test, FMT preceded by vancomycin pre-treatment may be offered again. It remains unclear whether an effect may be achieved by changing the donor or the application method.

Long-term follow-up

The long-term consequences of FMT are unknown. Treatment should therefore be performed with registration that allows for long-term follow-up. The attending physician plans and handles follow-up. The working group recommends a minimum one-year follow-up period.

Side effects and complications

All patients offered FMT should be informed in person and in writing about any risks of the treatment. The patient’s consent to treatment is oral, but follow-up for quality monitoring purposes after the patient’s course has been completed and for dissemination of health information can take place only after prior written patient consent. Adverse events and serious adverse events are documented by the attending physician. The individual FMT service defines how serious adverse events are registered and evaluated. Immediate side effects to FMT are seen in approximately one-third of patients [19]or more, as systematic reviews have found that adverse events after FMT are underreported [120,121]. Adverse events include abdominal pain, diarrhoea and rumbling. The patient should be informed of this prior to treatment. Long-term side effects are consequences that extend beyond the first 1–3 days after treatment. Serious adverse events are recorded in the patient record. When describing serious adverse events, the severity of the events and the causality of FMT are described [122].

Special situations

CDI in patients with inflammatory bowel disease

For treatment of CDI with FMT in patients with IBD, the focus has been on whether the effect of FMT is less in IBD patients than non-IBD patients, whether there is a risk of exacerbation of IBD after FMT and whether IBD disease activity has an effect on the effect of FMT. FMT among IBD patients with CDI has been elucidated in systematic reviews and meta-analyses of cohort studies [123,124] and in comparative studies of IBD patients versus non-IBD patients [125,126]. Several of the considered studies included relatively few patients, and the largest cohort studies included 67 patients with IBD. Overall, although evidence that FMT is effective among patients with IBD and concomitant CDI varied, no difference in treatment efficacy between IBD and non-IBD patients was reported [127–129]. Moreover, a comparable recurrence rate of CDI in IBD and non-IBD patients indicates that there is a similar effect in Crohn’s disease and ulcerative colitis. In a meta-analysis of 29 studies, worsening of IBD was found in 23% after FMT for CDI, whereas the risk of worsening was 11% in patients treated with FMT for IBD [124]. Treatment was also found to be cost-effective compared with vancomycin and fidaxomicin [130].

Severe immune incompetence

Treatment of patients with severe immune incompetence places particular demands on the balance of risk and benefit of FMT, as severe immune incompetence involves a theoretical risk of bacterial translocation following FMT. Two cohort studies [123,131] and a systematic review [132] have found that FMT used in patients with severe immune incompetence does not increase the incidence of complications [131], and treatment can generally be recommended [132,133]. Possible complications include bacteraemia, and the transmission of multi-resistant microorganisms is sought avoided through strict donor screening [134]. Despite thorough donor screening, the increased risk of transmission of potential unknown pathogens should be considered and the patient should be informed about this. As a definition of severe immune incompetence, the following is proposed: Current or predicted severe neutropenia (<0.5 billion/L) OR planned or recent (<100 days) allogeneic stem cell transplantation OR active graft-versus-host disease (GVHD) in need of immunosuppressive therapy. The significance of less pronounced degrees of immune incompetence for outcomes and complications after FMT remains unclear as is the case for patients with a reduced CD4 + cell count, patients on other immunosuppressive therapy such as biological therapy, azathioprine, methotrexate, nucleoside analogues, and patients with congenital or acquired primary immunodeficiency. In such patients, extended follow-up and monitoring may be planned. In patients with organ transplantation, FMT can be used, and a retrospective review does not indicate an increased incidence of complications in organ transplant recipients [123].

Liver cirrhosis

A retrospective cohort study found that patients with liver cirrhosis and CDI needed a larger dose of donor stool to achieve the same treatment effect as other patients with CDI. FMT can be used in patients with liver cirrhosis, but more treatments may be needed. The optimal treatment algorithm has not been clarified.

Condition with colectomy

Patients with ileostomy or ileorectal anastomosis may develop CDI, and the effect of FMT is case sensitive and should be considered [135].

Clinical recommendations

Clinical recommendations for the use of FMT are presented in Table 4.

Table 4. Clinical recommendations for the use of faecal microbiota transplantation (FMT).

Clinical recommendations for the use of faecal microbiota transplantation (FMT).
Contraindications and precautions	In conditions other than CDI, FMT should only be offered as part of clinical trials (EL 5, RG D). There are no known absolute contraindications to FMT (EL 5, RG D).
Patient preparation before FMT	In case of disclosure of health information, written patient consent is required (legal requirements). Remediation of antibiotics and proton pump inhibitor is performed as part of FMT (EL 5, RG D). Vancomycin is given 4–10 days before FMT and discontinued 1–2 days before FMT (EL 1a, RG A). In case of intolerance to vancomycin, fidaxomicin may be used as pre-treatment (EL 5, RG D). No evidence exists for the use of FMT without pre-treatment with antibiotics (EL 5, RG D). Purification before FMT is performed according to the application form but hardly has an independent clinical effect in FMT (EL 3b, RG C).
Clinical application and follow-up	Colonoscopic administration of FMT has a higher effect than the other application forms, and rectal installation has a lower effect than the other application forms (EL 1a, RG A). In rectal installation, repeated/repetitive treatment after/within 7–14 days produces an overall effect that may be in line with the other forms of application (EL 2a, RG B). In patients with fulminant CDI, repeated/repetitive treatment after 3–5 days may increase the clinical effect (EL 2a, RG B). In FMT for rCDI, the faecal component corresponds to 50 grams of stool from one donor (EL 2a, RG B). Patients who have received FMT for CDI are observed for 30 minutes after colonoscopic FMT in a right-sided position, for 1 hour in a left-sided position after rectal infusion, and sitting for 30 minutes after upper application (EL 5, RG D). The primary endpoint of FMT for CDI is the absence of CDI defined as normal bowel movements (three or fewer bowel movements per day), or in case of diarrhoea (more than three thin (Bristol 6–7) bowel movements daily) (EL 5, RG D). Long-term follow-up of all patients who have received FMT is generally performed for at least one year (EL 5, RG D). In patients who have recurrence after previous FMT, new FMT preceded by vancomycin may be offered (EL 5, RG D). Adverse reactions and complications are recorded (EL 5, RG D).
Special situations	In patients with known chronic inflammatory bowel disease and CDI, FMT can be used for common indications, but the patient must be informed of the risk of flare-up (EL 3b, RG C). In patients with severe immune incompetence, FMT can be used, but special caution and close follow-up need to be agreed (EL 3b, RG C). Organ transplants and other patients undergoing immunosuppressive therapy may be offered FMT in line with other patients (EL 3b, RG C). Patients with liver cirrhosis may be offered FMT on the usual indication, but additional treatments may be needed (EL 3b, RG C). In patients who are colectomized and develop CDI, FMT may be considered in the same way as in patients with preserved colon (EL 4, RG D).

Organisation of FMT, donor recruitment and screening, laboratory preparation

Organisation of an FMT service

Hospital-based non-commercial use

The principle of this guideline is that faecal suspensions are not subject to sale or marketing and are used only in hospital settings on established indications or as part of clinical trials.

Infrastructure

Routine use of FMT takes place as part of a complex treatment that involves both clinical and paraclinical specialties. A complete FMT service includes a formalised system, which can generally be divided into three domains: (1) donor recruitment and donor screening, (2) laboratory preparation and storage of faeces, and (3) clinical application, follow-up and management of complications. The first two domains may be organised in a centralised stool bank that handles the donor-related matters and preparation of donor faeces. A description of the responsible staff should be in place, including a description specifyingwho holds the paraclinical and clinical responsibilities, and the distribution of areas of work and responsibilities. Treatment may take place at the clinical ward of the stool bank or at a separate clinical ward (treatment ward) in collaboration with the stool bank. Treatment that takes place at a local hospital following supply of faeces from a stool bank should take place as part of a formalised collaboration agreement with and clinical medical support provided by the faeces bank’s staff. Clinical responsibility for the application of FMT lies with the attending physician. A clinical department where FMT is performed should have access to medical clinical, clinical immunological and clinical microbiological expertise in order to deal with issues and complications. Treatment can be done according to consensus-based, published protocols [136–139] and international guidelines [140,141]. General principles for donor recruitment, screening, preparation, release and clinical application are described in the Danish Tissue Act with accompanying technical guidance.

Regulation

The general structure of a stool bank and a clinical treatment department may be established according to the principles described in the Council of Europe Guideline on the Safety and Quality of Tissues and Cells for human application [140]. At the European level, however, the EU Commission currently finds that donor faeces is currently not covered by the Tissue and Cell Directive [142] and leaves it up to individual Member States to decide whether treatment should be regulated under this Directive [143] or in pursuance of the Medicines Act [144]. A process is underway to incorporate and classify donor faeces in the most appropriate legislation. Until FMT-specific legislation is available, the working group finds that the most suitable material is the standards established in the Danish Tissue Act for safe handling of donated human material, which ensures donor as well as recipient and which can be applied to donor faeces. The Tissues Act’s standards for selection of donors, donor screening, preparation and storage of tissue suspensions as well as general quality assurance are already established practice in all blood banks. These standards should guide the general structure of an FMT service’s donor handling and donation. In addition, the Council of Europe’s guideline on the safety and quality of human use of tissues and cells contains a separate chapter on FMT, which may further guide clinical management [140].

Documentation and quality assurance

The Danish Tissue Act, derived from the European Tissue and Cells directive [143], describes principles for establishing donor anonymity, traceability, storage of samples and other documentation. Guidelines for medical records are described in the Record Keeping Order (https://www.retsinformation.dk/Forms/R0710.aspx?id=201378). Registration of patients’ health information in the database may take place following registration of the data-responsible authority’s project overview database and requires written patient consent. Documentation of clinical activity and registration of adverse events may be done in accordance with the recommendations provided below. Examples of patient guidance, information material and a consent statement are available as appendices.

Donor recruitment

Faeces donation is voluntary and unpaid in accordance with the Danish Tissue Act. This increases the security of the donation material, as no financial incentive exists to donate; this also ensures that human material does not become the subject of trade, which would violate ethical standards [140,145].

Recruitment

Faeces donors may be recruited through promotion in blood banks or by personal contact. The general population’s suitability as a faeces donor is around 2–10% [146–148]. In order to achieve a cost-effective donor recruitment with the highest possible security and to ensure supply stability, recruitment via the public blood banks is recommended [139,147–153].

Anonymity

Faeces donors must be ensured anonymity in accordance with the Danish Tissue Act. This means that no dedicated donors are used, i.e. donors remain unknown to the recipient. It is recommended to archive and store all documents related to traceability for a minimum of 30 years in accordance with the Danish Tissue Act [143].

Donor consent

Prior to assessment as a faeces donor, the candidate donor should be informed in person and in writing about the course and about any potential risks. The candidate donor should also be examined for any carrier status of multi-resistant microorganisms. If conditions arise during the screening that require further investigation and/or treatment, the donor will be referred hereto. The responsible physician of the stool bank or FMT service is responsible for this. No inconveniences are associated with donating faeces. As personally identifiable data are being stored about the donor, written consent to the registration of health information in the database is required in accordance with applicable law. An agreement is reached on screening, number of donations and possible participation in a research project. An example of a declaration of consent is given in the appendix.

Donor screening

Donor screening includes pre-screening based on general criteria, detailed questionnaires and analysis of blood and faecal samples before commencing and after completingthe donation round (Table 5) [107,137–139,148,151,153]. The donor screening program is updated continuously, and exchange of experience through national and international networks may be expedient.

Table 5. Donor screening in faecal microbiota transplantation (FMT).

Pre-screening	Age 18–65 years Body Mass Index (BMI) 18–28 kg/m^2 Without/No allergy or chronic disease, without fixed medication consumption
Questionnaire	Without/No risk factors as with blood donation, according to the Tissue Act Without risk factors for MRSA and CPO Without current, chronic or previous oncological diseases Without intestinal symptoms (diarrhoea, constipation, bloody stools, abdominal pain, nausea, flatulence) Without previous bowel surgery (phased removal of appendix) Without depression or psychiatric disorder that requires medication Without hereditary incidence of bowel cancer, inflammatory bowel disease, celiac disease 0-3 months: No antibiotic treatment 0-6 months: No risk behaviour, blood transfusion, acute intestinal disease, travel history outside Northern Europe
Blood tests	Creatinine, alanine aminotransferase, haemoglobin, albumin, C-reactive protein. CMV (anti-CMV IgM/IgG), EBV (VCA-IgM, VCA-IgG and EBNA IgG), HAV IgM, HBV (HBs Ag, HBc Ab), HCV (HCV Ab), HIV (HIV1-2 Ab/Ag), Syphilis Ab and NAT test (HBV, HCV, HIV)
Stool samples	Intestinal pathogenic bacteria • Clostridioides difficile • Salmonella • Shigella • Campylobacter jejuni/coli • Yersinia enterocolitica • Diarrhoea-causing Escherichia coli: o Verotoxin-producing E. coli (VTEC), o Enteropathogenic E. coli (EPEC), o Enterotoxigen E. coli (ETEC), o Enteroinvasive E. coli (EIEC) o Intimin-producing E. coli (attaching and effacing E. coli (A/EEC)) Intestinal pathogenic viruses: • Adenovirus, astrovirus, sapovirus, enterovirus, parechovirus, norovirus, rotavirus Intestinal pathogenic parasites: • Entamoeba histolytica • Cryptosporidium parvum/hominis • Giardia lamblia • Worms Multi-resistant microorganisms: • ESBL-producing E. coli, Klebsiella pneumoniae and Proteus mirabilis • Carbapenemase-producing organisms (CPO) • Vancomycin-resistant enterococci (VRE) Others • Helicobacter pylori antigen • SARS-CoV-2 RNA

Approval of faeces donors

The overall review of donor screening is handled by a physician with the necessary qualifications who is affiliated with the stool bank or equivalent. Quarantine provisions following travel abroad, blood transfusion, antibiotic treatment, etc. are applied in accordance with the principles in the Danish Transfusion Medicine Standards (https://dski.dk/gaeldende-version-rejse/).

Stool donation

Donation of faeces takes place in donation rounds, which can last up to 30 days. Before the start and after the conclusion of any donation period, donor screening is completed, and any donated faeces are quarantined until the results of both screenings have been approved. The material can then be released for patient treatment.

Donation

Donor faeces are handed in by the donor in the stool bank or transported directly by a third party to the stool bank in sealed packaging. Each faecal donation should be accompanied by a written statement stating the donor time of donation, confirming that the donor has not been ill or exposed to infection since the latest donation and that the faeces were donated by the donor. Using a dedicated laboratory information system throughout the process, the donor’s statement will be an electronic questionnaire, developed specifically for faeces donors, similar to the questionnaire on attendance used for blood donation.

Storage during transport

During transport, donations must be stored in transport containers validated for transport of human material. Storage at room temperature for up to two hours does not appear to affect the bacterial composition [154]. Depending on local logistics, donations may be placed in a cooler bag with cooling elements during transport.

Personnel, physical environment, critical equipment, quality assurance and document management

In the stool bank, an organisation plan must be in place, all staff must have a job description and a competence management system must be in place.

The laboratory where preparation is performed should be established in connection with the specialties of clinical immunology or clinical microbiology to ensure sufficient professional competencies. Standard procedures, detailed laboratory instructions for all parts of faecal sample handling, manufacture and testing are developed and used. Document management is ensured by the faeces bank using version-controlled documents which are regularly reviewed as part of the general quality management system [136,139,155]. Critical equipment must be defined and its use must be logged.

Quality assurance is achieved in accordance with the general principles presented in the Council of Europe’s guide [140] and in the Danish Tissue Act with the accompanying executive order and guidance.

Preparation of donor faeces

Donor faeces are processed minimally and solely for the purpose of conserving and preparing the donation for clinical administration.

Preparation for donation

Prior to each donation, facilities and remedies such as scales and workstations must be cleaned and disinfected with a chlorine product (1,000 ppm), and all material (recyclable) cleaned and chlorinated-disinfected to avoid potential cross-contamination, including with spores. As an alternative to chlorine disinfection, sterilization may be used. All critical equipment used is disposable. In the absence of CE approval, used utensils should be approved for food use.

Duration of preparation

In most studies where FMT has been used, faecal preparation has been completed within six hours without loss of clinical effect. Preparation should begin no more than two hours after donation and be completed no more than six hours after donation. When storing donor faeces before preparation, donor faeces are stored at 4 °C [156].

Aerobic/anaerobic preparation

Although preservation of the intestinal anaerobic bacteria is theoretically desirable, special anaerobic technique in preparation is not documented as necessary for achieving a clinical response after FMT for rCDI [157,158], possibly owed to the high incidence of viable spores in the intestinal microbiota of healthy individuals [159].

Donor-specific preparation

Donor faeces are processed without mixing faeces from several donors, i.e. each faeces suspension for treatment consists of faeces from one donor. This is done to maintain traceability between donor and recipient and to ensure the concept of transfer of one single microbiota.

Homogenisation

Homogenisation may be done using a disposable stick blender, smasher, or stomacher bag [160]. No differences in clinical effect for degree or method of homogenization have been described.

Preparation for faecal suspension or capsules

Standardised protocols for preparation and freezing of donor faeces have been published [100,136,139] and validated with clinical endpoints for both liquid faecal suspensions [147], glycerol-preserved and encapsulated faeces [105,161] and freeze-dried donor faeces [115]. In preparation without freeze-drying, 10% glycerol is usually added to the donor faeces as a cryoprotectant [111,157]. The faeces suspension is homogenized in sterile isotonic sodium chloride [162]. None of the added adjuvants have therapeutic effects. For one treatment, 50 grams of donor stool is recommended [138].

Freezing of donor faeces

Cryopreservation is necessary for storing donor faeces until use. Frozen and fresh donor faeces have the same treatment effect [13,18,116,157]. Because post-donation donor screening is not possible with fresh donations, only frozen-thawed faeces should be used. In addition, this facilitates practical handling and reduces the risks associated with handling.

Archive samples (safety samples)

Archive samples are taken from all donations. Samples are stored at −80 °C for at least two years after the donor faeces is used.

Labelling of faeces suspensions

When marking and supplying suspensions to FMT, the stool bank must comply with the conditions described in the Danish Tissue Order (Executive Order no. 58 of 18/01/2019, Order on quality and safety in the handling of human tissues and cells, Appendix 2, Sections D and E and Annex 8, Section 1.6 and Section 1.7). All FMT suspensions must be marked at the time of removal. As a minimum, the primary container must be marked with donation identification data/code and tissue type.

The container with the FMT suspension must be marked with:

• Tissue type, tissue identification number/code and lot or batch number, tissue centre identification and expiration date.

• For recipient-specific donations, the label must state for whom the donation is intended.

• FMT suspensions must be labelled with a common European code (SEC), e.g. the ISBT 128 coding system, if the suspension is distributed for human use by another legal entity (hospital unit) other than the tissue centre.

If no space is available for the information on the container’s label, it must be indicated on a separate delivery note accompanying the container. The delivery note must be packed together with the container in a manner that ensures that they remain together at all times. In addition to the above, the label or delivery note must contain the following information:

• Description (definition) and, if applicable, the size of the tissue or cell product

• Date of tissue distribution

• Selected biological determinations performed on the donor and the results

• Recommendations regarding storage

• Instructions for opening the container, packing and any handling/reconstitution required

• Shelf life after opening/handling

• Instructions for reporting serious side effects and adverse events

When supplying FMT suspensions, the procedures of the approved tissue centre must meet the following criteria:

• Critical transport conditions, e.g. temperature and time limit, are determined so that the required tissue and cell properties are maintained

• The container/package must be secure and ensure that the tissue and cells are preserved under the conditions specified. All containers and packages must be validated as being fit for purpose

• If FMT suspensions are distributed by a third party under contract, a documented agreement ensuring that the required conditions are present must be produced

• Tissue establishments must have approved staff who can assess the need for recall and implement and coordinate the necessary measures

• An effective recall procedure must be in place that includes a description of the division of responsibilities and measures to be taken. The procedure must include notification to the Danish Agency for Patient Safety

• A documented system must be in place for handling returned products, including criteria for including them in the inventory.

Storage, quarantine, release and distribution

Storage

Stools are stored suspended in 10% glycerol at −80 °C; tissue suspensions must be stored at −80 °C or cooler. After storage, clinical effect has been documented for 6–10 months, and shelf life may be longer [163]. Storage at −20 °C is possible for up to one month [164].

Quarantine

After preparation, stool suspensions must be placed in a separate section for quarantined suspensions until all release criteria are met. There should be clear markings and a control system to handle and separate released faecal suspensions from quarantined ones. Marking follows local instructions or standard operating procedures (SOP).

Release of donor faeces

Before the FMT suspension can be released, ensure that: (a) The donor is suitably assessed on the basis of initial donor screening. (b) A signed consent form exists. (c) Signed delivery notes from all faecal donations are available. (d) All analysis results are available and well known, both from the initial screening and from the first and last donation within the donation round. (e) The FMT suspension was processed and frozen no later than six hours after the donation time.

Distribution

As the use of FMT is developed and established as a clinical treatment option, simultaneous treatment is made possible locally in specialist departments by qualified specialists. The fact that the treatment can be performed locally has benefits for patients and therapists alike. It also places demands on logistics, quality assurance and documentation at the local treatment centre. Distribution may take place as part of a formalised collaboration between the stool bank and the treatment site. The conditions needed to establish FMT at the clinical treatment department are established in dialogue with the distributing stool bank. As the treatment principle is under development, any application is subject to registration and annual evaluation of clinical activity, effect, adverse events, etc. [165].

Preparation before treatment

The day before its clinical use, donor faeces stored in cryotubes, cryobags, or capsules may be transferred to −20 °C. Generally, storage at −20 °C is acceptable for up to one month [163]. During distribution to another location donor faeces is stored on dry ice. Defrosting is done according to local instructions. Thawed donor faeces should be used within two hours and should not be re-frozen, as repeated freeze-thaw cycles may reduce viability [166].

Clinical application and safety parameters

Clinical application follows the general descriptions above. Clinical responsibility lies with the treating specialist.

Deviations

There must be local procedures for reporting deviations (e.g. risk of infection or possible unwanted side effects) to the physician in charge. Stool banks must ensure communication channels before commencing faeces distribution to clinical departments and the clinical departments must be able to handle any clinical complications.

Traceability

Traceability must be ensured for each FMT suspension. Traceability includes unique identification allocated to each donation, traceability between donor and suspension, suspension records, suspension status and recipient, as well as critical equipment used in preparation. Stool banks must register data on the patient for whom the FMT suspension is handed out for treatment so as to ensure traceability between donor and recipient.

Look-back

If it is suspected that a patient may have become infected with pathogenic microorganisms through FMT, look-back should be initiated. An archive sample (described above) is stored from each faecal donation, which is used for look-back and, if relevant, for testing for a specific pathogen.

Follow-up and registration of adverse events must be done as described above.

Internal audits must be conducted in accordance with the Danish Tissue Act, Appendix 1, Section F. Internal audits must be conducted at least every two years in order to verify that the approved internal instructions (SOPs) are complied with.

Minimum standards

Suggestion for minimum standards for implementing an FMT service are presented in Table 6.

Table 6. Mininum standards for implementing a faecal microbiota transplantation (FMT) service.

Mininum standards for implementing a faecal microbiota transplantation (FMT) service

An FMT service has a defined organisation, including a description of the persons involved and the division of responsibilities between them. An FMT service complies with the principles of the Tissue Act, including principles for traceability and anonymity as well as documentation of clinical activity and safety parameters. An FMT service has ensured access to qualified medical conference and development. Faecal donation is voluntary and unpaid. Donor recruitment, donor screening and faecal donation are established and managed in public blood banks. Faeces donors provide their written consent for the registration of personally identifiable data in the database. Donor screening includes clinical information, blood tests and stool samples with an updated donor screening panel before or in connection with any commenced and completed donation round. Laboratory processing follows published protocols and guidelines and is performed by specially trained personnel. The processing of each faecal donation is documented. Archive samples from faecal donations are stored for at least two years. Critical equipment and critical processes are defined and documented. Release of FMT components presupposes that the overall donor screening has been approved, that the donor has signed consent declarations and donation delivery notes and that both declarations and notes have been processed correctly. Distribution takes place according to written cooperation agreements in which the parties' tasks and responsibilities are described. FMT is used within the framework agreed with the relevant health authorities. Traceability between donor and recipient as well as full anonymity are ensured. Documentation regarding traceability is stored for 30 years. Clinical trials are monitored by a local GCP unit. Clinical application takes place in the relevant medical specialty, which handles patient assessment, clinical application, follow-up on and management of any complications. Critical events are recorded and evaluated.

The emergence of societal and health-related patterns, exemplified by the recent COVID-19 pandemic, will likely require a dynamic adaptation of stool bank and FMT services in the future [167].

Abbreviations
CD	=	Clostridioides difficile
EL	=	evidence level
FMT	=	faecal microbiota transplantation
IBS	=	irritable bowel syndrome
NNT	=	number needed to treat
OVP	=	oral vancomycin prophylaxis
RBI	=	rectal bacterial instillation
RBT	=	rectal bacteriotherapy
rCDI	=	recurrent Clostridioides difficile infection
RCT	=	randomised clinical trial
RG	=	recommendation grade
SOP	=	standard operating procedure.

Acknowledgements

This Danish national guideline was endorsed by the Danish Society for Gastroenterology and Hepatology, the Danish Society of Infectious Diseases, the Danish Society for Clinical Microbiology, and the Danish Immunology Society for Clinical Immunology.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Innovation Fund Denmark (j.no. 8056-00006B) supported with funding through the project Centre for faecal microbiota transplantation at Aarhus University Hospital (CEFTA). The funder had no influence on the content of the manuscript.